# Allocate Resources

> >

- **Type:** Skill
- **Install:** `agentstack add skill-daddia-claude-for-strategy-allocate-resources`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [daddia](https://agentstack.voostack.com/s/daddia)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [daddia](https://github.com/daddia)
- **Source:** https://github.com/daddia/claude-for-strategy/tree/main/corporate-strategy/skills/allocate-resources

## Install

```sh
agentstack add skill-daddia-claude-for-strategy-allocate-resources
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Allocate Resources

## When to use

For CSO, portfolio managers, or strategy leads diagnosing whether capital, headcount, and management attention match stated strategic priority — not history, politics, or sunk cost.

## What this skill does not do

- **Does not infer priority from current spend** — that's the failure mode being diagnosed.
- **Does not invent budget or headcount** — `INPUT NEEDED` when data missing.
- **Does not approve reallocation** — recommends specific from→to moves for portfolio forum.
- **Does not run growth-vector assessment** — cross-checks `assess-growth-vectors` if available.

## Preconditions

| Input | If missing |
|---|---|
| Org + corporate-strategy profiles | Surface practice-setup or `[PROVISIONAL]` |
| Strategic priority ranking | Halt — get or confirm before comparing |
| Allocation data (capital, HC, attention) | Ask; mark partial/missing per row |

## Provisional mode

Defaults: generic portfolio units; three priority tiers; `[PROVISIONAL]` on all figures. User must provide dollars/headcount — never invented.

## Trust spine

- **Confidence bands** (`option-evaluation`): High = all options scored against same priority framework, sourced; Medium = criteria explicit, some data partial; Low = priority or spend unknown — structured first pass only.
- **Strategic advice vs. support (mandatory):** Reallocation table is recommendation input — `[review]` on each move; strategist and portfolio committee decide.
- **Escalation:** Missing priority → stop; inferring priority from spend → forbidden; board-final → GATE + reviewer note.

```
SOURCING / ASSUMPTIONS / NUMBERS / CONFIDENCE / GATE — per trust-conventions
```

## Workflow

### Step 1: Orient

Portfolio scope, allocation types, review forum, time horizon, evidence pack. Read profile definitions for units, tiers, stickiness.

### Step 2: Evidence pack

Capital/budget, headcount, management attention, priority ranking — status and source per row. Management attention: `[user estimate — not verified]` if unsourced. Figures from `~~bi analytics`, `~~spreadsheet`, or `~~crm` connectors must carry `[sourced: , , ]` on the load-bearing digit.

### Step 3: Allocation vs priority comparison

Per unit: priority tier/rank alongside capital %, headcount %, attention. Tag `[sourced]` or `INPUT NEEDED`.

### Step 4: Pattern detection

| Pattern | Flag when |
|---|---|
| **Legacy overfunding** | Low priority, disproportionate share |
| **Under-resourced growth bet** | Top-tier bet, bottom-quartile spend |
| **Under-differentiated allocation** | Flat spend despite wide priority spread |

Ask why for legacy overfunding — history is diagnosis, not justification.

### Step 5: Growth bets check

Cross-check `assess-growth-vectors` if run; else flag partial check.

### Step 6: Reallocation recommendation

Specific from→to moves with rationale — not vague "rebalance toward growth."

### Step 7: Portfolio committee gate

GATE before board/portfolio-committee finals.

## Output format

```
CONFIDENCE: [defensible recommendation | structured first pass]
SCOPE: [units]
REVIEW FORUM: [working | portfolio committee | board]

LOAD-BEARING ASSUMPTIONS: [...]

ALLOCATION VIEW:
| Unit | Priority | Capital % | Headcount % | Attention | Flag |
...

UNDER-DIFFERENTIATION CHECK: [...]
GROWTH BETS CHECK: [...]

RECOMMENDED REALLOCATION:
| From | To | Type | Magnitude | Rationale |
...

STICKINESS NOTE: [...]
EVIDENCE GAPS: [...]
```

## Worked example

**Input:** Unit A priority tier 3 (exit candidate), 28% capital share; Unit B tier 1 adjacent bet, 8% capital.

**Excerpt:** Unit A flagged legacy overfunding; Unit B under-resourced growth bet. Reallocation: A −10% capital → B +10% `[review]` — adjacent bet cannot scale at current funding.

## Quality checks before delivering

- [ ] Profiles loaded or `[PROVISIONAL]`
- [ ] Priority not inferred from spend
- [ ] No invented figures
- [ ] Patterns named explicitly
- [ ] Reallocation specific from→to
- [ ] Growth bets checked where data exists

## Propose profile update

When a stable convention surfaces during this run (thresholds, naming, tone, output format, or recurring corrections), **propose a profile update**: show the exact diff against `~/.claude/plugins/config/claude-for-strategy/corporate-strategy/CLAUDE.md` (org-wide facts go to `org-profile.md`), ask for confirmation, and write only on yes. Only `/corporate-strategy:practice-setup` auto-applies a full profile write.

## Outputs

Follows plugin `CLAUDE.md` § Outputs. Branches: supply missing data, portfolio committee memo, `exit-or-double-down` on legacy units, `assess-growth-vectors`, or set reallocation trigger date.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [daddia](https://github.com/daddia)
- **Source:** [daddia/claude-for-strategy](https://github.com/daddia/claude-for-strategy)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-daddia-claude-for-strategy-allocate-resources
- Seller: https://agentstack.voostack.com/s/daddia
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.